My work with FrontRow

The first question I ask is not about the product. It is: what do these people already believe about this person?

A cricketer's followers do not arrive wanting a curriculum. They arrive wanting proximity. A rapper's followers arrive wanting to sound like him. Those are two different products with the same price, and a page that treats them as one thing converts like an advertisement.

Three things follow from that, and they are the order I would work in again.

Lock the window before you write anything. When the collaborator is a public figure, their calendar, their legal review and their own posting schedule set the date, and the date will not move for you. Everything else is negotiable. That is the opposite of a normal launch, where the date is the last thing fixed.

Build the page out of their words, not about them. The lines that worked were things the person had said in their own register, lifted whole. Marketing copy written about a famous person reads as borrowed, and an audience that follows them can hear the difference in one sentence.

Treat the launch as a template on the first one, not the fourth. The temptation with a well-known collaborator is to treat every launch as unrepeatable. Nine of the ten decisions repeat: the page structure, the creative test, the sequence after signup, the way the offer is framed. Only the person, the craft and the dates change. If each launch starts from a blank page, the team spends its whole capacity on production and never on the thing that is actually losing money.

What usually turns out to be wrong is the join. The ad promises the person and the product delivers a course. Attention is cheap when a famous name does the asking, and attention does not convert into patience.

What goes wrong between the free lesson and the first payment?

This is where I learned the most, and it is the step most companies do not measure separately.

Count four numbers, not one: how many started the free lesson, how many finished it, how many opened the paid page, and how many paid. The drop is almost always between finished and opened. That is a product and copy problem wearing the costume of a traffic problem, and no amount of extra spend touches it.

Then ask what the free lesson ends on. If it ends on a summary, the person is finished. If it ends on something they can go and try that evening, they come back to ask the next question, and the next question is what the paid course answers. The end of the free thing is the beginning of the paid thing. I have used that in every consumer product I have worked on since.

The two variables nobody wants to test here are price and length, because they feel like product decisions rather than growth decisions. They are the offer. A shorter course at a higher price and a longer course at a lower price are different businesses, and you can find out which one you have in a few weeks if you are willing to run the test before the polish.

What should the messages after signup do?

One decision per message, in the order a person actually reaches it.

The common failure is a welcome email followed by a newsletter. That is a broadcast schedule, not a sequence, and it is why so much lifecycle revenue sits flat. For a course the decisions are legible: did you start, did you get stuck, what does the next lesson give you, what does paying get you, and what happens if you stop now.

Timing follows behavior rather than the calendar. A message on day three to someone who has not opened the app at all is a different message from one to someone who finished two lessons on day one, and sending both people the same thing teaches you nothing about either. If I could keep only one message in a sequence like this, it would be the one that goes when the second session does not happen. The first session is enthusiasm. The second is a decision.

The full shape, five messages and what each one has to answer, is in Lifecycle emails that aren't a newsletter.

Who built FrontRow, and who backed it?

Three founders. Ishaan Preet Singh studied at IIT Delhi and worked at McKinsey and at Lightspeed, where he is now an investor again. Mikhil Raj also studied at IIT Delhi; his first startup, PhotographAI, was bought by ZestMoney. Shubhadit Sharma was VP Engineering at GlowRoad. Ishaan and Mikhil were on Forbes India's 30 Under 30 in 2021.

FrontRow raised $17.2M in two rounds. The $3.2M seed in 2020 was co-led by Lightspeed and Elevation Capital, with Deepika Padukone's family office. The $14M Series A in 2021 was led by Eight Roads Ventures and GSV, with angels including Kunal Shah of CRED, Gaurav Munjal of Unacademy, Vishal Dadlani and Raftaar.

Lightspeed has also backed Anthropic, Snap and Databricks, and Razorpay, OYO and Zepto in India. Elevation Capital has also backed Paytm, Swiggy, Meesho and Urban Company.

What were FrontRow's own numbers?

By its co-founder's account, FrontRow reached 2 million downloads and more than 250,000 people who paid for a course.

FrontRow closed in 2023, and its co-founder wrote openly about why. Reading that taught me as much as the launches did, and it is the reason I treat a flat second-week number as an emergency rather than a reporting detail.